Solicit and incorporate final feedback on server side CSV generation

[danyx23]

Now that #3613 (and some follow-up PRs) have been merged, the basic feature of downloading csv files from a URL works - but the ergonomics of the server side csv could do with a final round of polish. To this end, interview, triage and incorporate the feedback from

• Our data managers
• Lars and the other devs
  • Lars already provided some immediate feedback in this slack thread
• Joe (especially regarding the contents of the zip file and the importantce of an Excel file)
• Este (especially regarding the wording in the readme file that touches licensing-ish questions)